ATLANTA (GenomeWeb) – Data from two studies presented at the annual meeting of the American Association for Cancer Research on Sunday highlighted researchers' efforts to design safe and effective chimeric antigen receptor T cells (CAR T) against solid tumors by targeting proteins that tend to be expressed more on cancer cells than normal cells.
The FDA has approved two CAR T cell therapies to date, tisagenlecleucel (Kymriah) for the treatment of young adults with acute lymphoblastic leukemia and axicabtagene ciloleucel (Yescarta) for the treatment of children and adults with diffuse large B-cell lymphoma. Both of these CAR T therapies target CD19, which are widely present in B-cell malignancies and a subset of leukemias, but aren't critical to the survival of normal cells.
Following the success of CAR T therapy in treating these blood cancers, researchers are turning their attention to the more difficult task of engineering these immunotherapeutics to recognize and attack solid tumors. Unlike in blood cancers, few antigens have been identified in solid tumors that are expressed intensely on cancer cells but expressed at low levels or absent on normal cells.
Researchers have been testing out several tumor surface proteins, and HER2 and mesothelin are two that have garnered significant interest. One Phase I study, led by Prasad Adusumilli, deputy chief of the thoracic service at Memorial Sloan Kettering Cancer Center, enrolled 21 patients with mesothelioma of the chest cavity, or breast and lung cancer that has spread to this region, and who had tumors expressing a protein called mesothelin. His team treated patients with CAR T cells engineered to not only target this protein but also with a "suicide switch" that can be turned on to destroy the CAR T cells in the event of toxicity.
Some of the patients in this study who experienced complete or partial responses also received treatment with an anti-PD-1 inhibitor, which demonstrates the potential efficacy of this combination strategy. At the meeting, Nilofer Azad from Johns Hopkins University's Sidney Kimmel Comprehensive Cancer Center, said that "these data are potentially the most compelling CAR T data we've ever seen in solid tumors at this point."
Another early-stage study, led by Meenakshi Hegde from the Center for Cell and Gene Therapy at Baylor College of Medicine, treated 10 sarcoma patients with CAR T cells designed to target HER2 receptors on the tumor surface. "HER2 is expressed in very limited ways in normal tissue and is upregulated or aberrantly expressed only in cancers," said Shoba Navai, assistant professor of pediatrics at Baylor's Center for Cell and Gene Therapy, who was a co-investigator on the study. "HER2 is expressed in a variety of sarcomas, and because of the limited tumor antigens available to target in solid tumors, this was the starting place for our research group many years ago."
When T cells reach the tumor, they look for the target antigens but can fail to identify them if the antigen isn't strongly expressed on the cell surface.
For this reason, in the study involving HER2-targeting CAR T cells sarcoma patients had to be tested for HER2 expression in order to be eligible for the study. Any detectable HER2 expression as determined by an independent pathologist (minimum grade 1, intensity 1+) was deemed sufficient for study enrollment. Around 40 percent of patients with a common type of sarcoma, called osteosarcoma, express HER2, but currently it's not well known how prevalent HER2 expression is in other types of sarcomas.
Even when osteosarcoma patients express HER2, it's at a level too low to be targeted by monoclonal antibodies, such as trastuzumab (Herceptin). Researchers from Baylor College of Medicine have previously shown in in vitro and in mouse models that HER2-targeted CAR T cells can target low levels of HER2 expressed on cancer cells when HER2 antibodies, such as trastuzumab (Herceptin), can't.
In the study by Adusumilli and colleagues, the antigen of interest is mesothelin, a protein that is expressed in a majority of the solid tumors occurring in an estimated 2 million cancer patients in the US per year. In the lab, researchers tested 2,000 tumor samples and normal tissue samples to quantify mesothelin. "We took that into consideration in designing the CAR T cells so they will go after the cells with high and moderate expression but not very low expression," he told GenomeWeb.
Researchers enrolled patients whose tumors expressed mesothelin or if they had high mesothelin levels in blood. "The reason we decided to target mesothelin is not because it occurs in 2 million patients," Adusimilli said, noting that mesothelin-expressing cancers tend to be very aggressive. "If the cancer cell doesn't need what I'm targeting, it's going to shed it. I want to make sure the cancer cell needs what I'm targeting."
However, in this study, researchers also included patients with lower mesothelin expression levels on tumors, and one of the patients who had a complete metabolic response to CAR T cell therapy for 16 months had mesothelin expression in 30 percent to 40 percent of tumor cells. Adusumilli predicted that the likely reason for this may be because the patient received lymphodepleting chemotherapy to allow for CAR T cells to expand and then received anti-PD-1 treatment several weeks into the infusions to overcome immunosuppression. In this way, researchers were able to "tilt the balance toward more effector-immune response" in this and other responding patients.
In the real world, patients will need to be tested for mesothelin expression in order to receive this CAR T therapy, but in patients with tumors that have spread to multiple sites, the challenge, Adusumilli said, is figuring out which tumor to test for mesothelin. "These are the real problems we face," he said.
Both of these studies involved tumor types with limited treatment options, which made the early responses seen with CAR T therapies in these trials encouraging. Sarcoma patients with relapsed or refractory disease have not had improvements in survival over the last four decades. The last treatment approved by the US Food and Drug Administration for malignant pleural mesothelioma was in 2003, which showed an overall survival benefit of around three months in clinical trials. These patients also don't respond that well to anti-PD-1/PD-L1 immunotherapy, but treatment guidelines recommend using anti-PD1 drugs as a second-line option.
Targeting mesothelin
In the US, about 150,000 patients a year have mesothelioma or breast and lung cancer that spread to the pleural cavity. Preclinical studies showed that when CAR T cells were injected intravenously in mice, they were being sequestered in the lung for four to five days and then trickling out to the tumor. This adds to the challenge of getting CAR T cells to the center of the solid tumor masses.
Adusumilli's group wanted to avoid lung sequestration and injected the CAR T cells directly into the pleural cavity. In prior studies, researchers have demonstrated that using this strategy, within a few hours of injection the T cells started circulating in the blood, and because they were injected regionally to the chest cavity, they seemed to recognize the antigen more readily, get activated, and start proliferating. "This is one of the reasons we're seeing the efficacy," he said.
In the study led by Adusumilli, researchers enrolled 21 patients — 19 with pleural mesothelioma, one with metastatic lung cancer, and one with metastatic breast cancer — and injected the CAR T cells directly into the chest cavity using an interventional radiology procedure.
Two patients had a complete response, eight had a partial response, two patients had stable disease, nine patients progressed. Researchers also selected a subset of 11 patients to receive lymphodepletion treatment first, then CAR T infusions, and after a period of six-to-eight weeks and making sure that these patients weren't experiencing serious side effects, gave them an anti-PD-1 drug, pembrolizumab (Keytruda).
One way tumors try to evade an immune attack is by upregulating expression of PD-L1, and this may be a reason why CAR T cells become exhausted in patients with large tumors. But prior experiments have shown that if patients are treated with an anti-PD-1 agent, it blocks immunosuppressive signaling and reactivates CAR T cells, allowing them to proliferate again and attack cancer cells.
Adusumilli's group used this approach, and eight of the 11 patients in this portion of the study had a complete or partial response, resulting in a response rate of 72 percent. This demonstrated that "we can rescue the functionally exhausted CAR T cells and prolong the tumor regressions," Adusumilli said.
Interestingly, four out of the eight responding patients were negative for PD-L1 expression. Two patients experienced complete responses for 16 weeks and another patient has experienced a 38-week complete response that's ongoing. Some patients who saw their tumors partially shrink responded for a year.
"This strongly supports pursuing a CAR T cell therapy combined with anti-PD-1 strategy in solid tumors," Adusumilli said. "Our strategy is to make these cold tumors hot by injecting CAR T cells first, then activating them functionally and rescuing them by keeping them warm."
None of the patients in this study experienced higher than Grade 2 neurotoxicities, cytokine release syndrome (characterized by high fever, heart dysfunction, liver failure, and kidney impairment), or on- or off-target toxicities at the administered doses. Some patients treated with anti-PD-1 agents experienced Grade 3 toxicities, but none of them hindered the ability of the patient to partake in the study.
Mesothelin is expressed in low levels in tissues around the heart, lungs, and abdominal cavity and so on-target, off-tumor toxicity was a concern. As a result, researchers designed a safety mechanism where in the event a patient experiences serious toxicities, they could treat the patient with a drug that activates a "suicide gene" that directs elimination of the CAR T cells within hours. In the study, however, because there weren't serious safety issues with the CAR T treatment, this switch didn't have to be used.
Next, Adusumilli's group will submit the data on the benefit seen with the addition of anti-PD-1 drugs to the FDA with the hope that the agency will allow researchers to give these drugs to patients earlier, at week three or four following CAR T cell therapy, instead of waiting six to eight weeks. But recognizing that this treatment comes with its own set of toxicities, his group has developed a new CAR T cell with a decoy PD-1 receptor that can bind to PD-L1 on tumor cells but won't get inhibited.
Researchers hope to take this CAR T cell therapy into clinical trials next year. If this is effective, "we don't have to keep giving anti-PD-1 therapy to the patient every three weeks for the rest of their lives, and we can avoid toxicities from checkpoint blockade," Adusumilli said. The technology related to mesothelin-targeted CAR and PD-1 dominant negative receptor has been licensed to Atara Biotherapeutics.
Targeting HER2
In the study involving HER2-targeted CAR T therapy, 10 patients ranging in age from 2 years to 54 years had received as many as five salvage treatments for advanced HER2-positive sarcomas. Patients received lymphodepletion chemotherapy to clear out other T cells and allow the CAR T cells to proliferate. CAR T cells were detected in all patients at six weeks of follow up. After receiving up to three infusions of the CAR T therapy, two patients had a complete response, four had stable disease lasting for 3.4 to 9 months, and four patients progressed. Five patients were alive at the last follow-up period.
One 16-year-old patient with osteosarcoma who had surgical resection of a tumor that had spread to the lung received CAR T infusions and remains alive without recurrence for nearly 35 months. Another eight-year-old boy with rhabdomyosarcoma with metastasis to the bone marrow saw the tumor disappear for a year. Upon relapse, this patient was retreated with HER2-targeted CAR T therapy, which led to another complete response that was still ongoing at 17 months.
Researchers tested this patient's serum for antibody responses against known human proteins and identified new antibodies after treatment, the majority of which were directed against non-HER2 intracellular proteins. "Many of these proteins are interconnected with pathways that are associated with promotion of cell proliferation, survival, and tumor metastases," Navai said. "These findings suggest that the patient's own immune system may be engaged to elicit new immune responses after treatment with HER2 CAR T cells and lymphodepletion chemotherapy." Researchers are exploring this in other patients.
One of the main challenges in designing CAR T cells in solid tumors has been to figure out how to ensure T cells reach the center of cancer masses and keeping the hostile environment around a tumor from diminishing their activity. The patients who progressed on the HER2-targeted CAR T therapy may be the result of them having large tumor burdens or having their cancer spread to difficult to treat regions.
"We think lung metastasis or bone marrow may be more favorable sites," Navai said. She added that as the tumors get bigger, some of the immune inhibitory signals coming from within the tumor may get more pronounced, making it more difficult to deliver CAR T cells.
All patients had low blood counts, which is expected with chemotherapy, but none developed infections or had decreased heart function known to be caused by HER2-targeted treatment, such as trastuzumab (Herceptin), or pulmonary toxicities that may occur when CAR T cells expand in the body. Some patients developed cytokine release syndrome, but these were not serious and were managed relatively quickly with supportive care.
In one prior study of trastuzumab-based HER2-targeted CAR T cell therapy, one patient had died from toxicities, but no serious events of this kind occurred in this study. Navai speculated that this may be due to the fact that the antibody portion of the CAR is not trastuzumab-based but uses a different HER2 antibody, FRP5.
The HER2-targeted CAR T trial will continue to recruit patients for the next year and test out the treatment in other HER2-expressing cancer. "We're also making modifications to our T cells to make them more efficacious," and target other elements in the tumor beyond HER2, Navai said. Similar to Adusumilli's group at MSK, Navai and colleagues also plan to study the combination of CAR T therapy with checkpoint inhibitors.